Stress and the Adrenocortical Control of Epinephrine Synthesis Richard J. Wurtman
Psychologic states produced by environmental or physiologic stresses are usually associated with hypersecretion of adrenal hormones. particularly epinephrine and the glucocorticoids (hydrocortisone in humans or corticosterone in rats). A common mechanism links the secretion of these hormones. even though the adrenal medulla and cortex have different embryologic origins and biochemical properties and very different mechanisms controlling their secretory activities. ie. a cholinergic nervous input stimulates medullary secretion while a hormone. corticotropin (ACTHI. activates secretion from the cortex. This mechanism is made possible by an intra-adrenal portal vascular system. which provides the medulla with uniquely high concentrations of glucocorticoids. These high concentrations are needed to induce the medullary enzyme. phenylethano- lamine-N-methyltransferase (PNMT). which controls the synthesis of epinephrine from norepinephrine. By suppressing glucocorticoid secretion. pituitary failure compromises epinephrine synthesis and decreases the rate at which epinephrine is secreted; in contrast. prolonged chronic stress can enhance epinephrine synthesis and secretion within the adrenal. the brain. or both organs. This control mechanism could be involved in the long-term consequences of stress. Copyright 2002. Elsevier Science (USA). All rights reserved.
HE CLASSIC PERIPHERAL manifestations of severe rod, to determine whether pituitary failure could be shown to T environmental or physiological stress-increased plasma affect the synthesis or secretion of this catecholamine. These levels of epinephrine and hydrocortisone (in humans) or corti- studies led to the discovery that the very high concentrations of costerone (in rats)-result from hypersecretion of these hor- glucocorticoids that the adrenal cortex normally delivers to the mones from the adrenal medulla and cortex, respectively. A adrenal medulla (via the adrenal's portal vascular system) in- common mechanism links these processes, even though the duce the enzyme, PNMT, that is rate-limiting in epinephrine medulla and cortex have different embryologic origins and synthesis. By suppressing glucocorticoid secretion, pituitary biochemical properties, and very different chemical signals (the failure not only compromised epinephrine synthesis, but also neurotransmitter acetylcholine and a pituitary hormone, corti- decreased the levels of epinephrine in the adrenal medulla and cotropin [ACTH), respectively) control their secretory activi- the rate at which epinephrine is secreted in response to hypo- ties. For several decades, on the basis of experiments per- glycemia.5 formed on rats, it had been thought that circulating epinephrine A severe prolonged stress, which chronically elevates glu- acted directly on the pituitary to control the release of ACTH. cocorticoid secretion, can have opposite effects. In this formulation the first step in the adrenal "stress response" ADRENAL MORPHOLOGY AND EPINEPHRINE SYNTHESIS was thought to be the liberation of epinephrine; this in turn released ACTH, leading to hypersecretion by the
Metabolism. Vol 51, No 6, Suppl 1 (June), 2002: pp 11-14 11
-- 12 RICHARD J. WURTMAN
Table 1. Effect of Hypophysectomy, ACTH, and Dexamethasone dulla in much lower concentrations than those available after on PNMT ACTH administration). T reatment* Adrenal Weight (mg/pairl PNMT (,..mol/pair) GLUCOCORTICOIDSAND EPINEPHRINESYNTHESIS Control 68.9 :!: 3.4t 6.16 :!: 0.66t Hypophysectomy 26.5 :!: 2.1 1.46:!: 0.11 There were at least 3 possible mechanisms by which ACTH Hypophysectomy plus treatment could have restored PNMT activity in the hypophy- ACTH 46.7 :!: 1.5t 4.76 :!: 0.26t sectomized rat: Hypophysectomy plus I. The adrenal cortex might have contained an unrecog- dexamethasone 26.0 :!: 2.3 7.80 :!: O.44t nized epinephrine-forming enzyme, in which case the *Animals were sham-operated or hypophysectomized and killed 17 decline in cortical mass after hypophysectomy, and its to 21 days later. Some of the hypophysectomized rats were given restoration with ACTH treatment, would be expected to ACTH or dexamethasone for 6 days prior to autopsy. cause parallel changes in PNMT activity. tp < .001 v hypophysectomy. 2. ACTH could have acted directly on the adrenal medulla to enhance PNMT activity. 3. ACTH could have stimulated the secretion of glucocor- In 1959, Kirshner identified the adrenomedullary enzyme, ticoids by the adrenal cortex; these compounds, deliv- PNMT, responsible for the synthesis of epinephrine from nor- ered to the medulla by the intra-adrenal portal vascular epinephrine.7 Subsequently, Axelrod demonstrated that most of system, could then have increased PNMT activity. the epinephrine-forming ability found in adult mammals was To examine the first possibility, we separated the cortex and localized within the adrenal glandS; hence, most of the epineph- medulla from adrenals of hypophysectomized and normal an- rine released into the circulation had very likely been produced imals, and assayed them individually for PNMT. We found that by PNMT within the adrenal medulla. An isotopic assay was the cortex normally contains less than 10% of the epinephrine- subsequently developed for PNMT, which simplified measur- forming activity of the medulla. (Since no capsule separates the ing the effects of various physiologic states and drugs on rat cortex from the medulla, it is possible that even this 10% epinephrine biosynthesis. represented contamination of-the cortic~1sample with medul- PITUITARYACTHAND EPINEPRINESYNTHESIS lary tissue.) To rule out the possibility that ACTH acted directly on the Based on the pathologically prolonged hypoglycemia of our adrenal medulla, it was necessary to show that when the level hypopituitary patient, we initiated studies to see whether the of glucocorticoids available to the medulla was held constant, glucocorticoid hormones were the unknown "methylation fac- changes in the rate of ACTH secretion had no effect on PNMT tor." Hypophysectomy was found to be associated with a activity}OThis was done in 2 ways. First, normal rats were marked reduction in epinephrine-forming ability (Table I)} treated with large doses of dexamethasone, a highly potent This reduction was of even greater magnitude than the fall in synthetic glucocorticoid. This dose depressed pituitary secre- adrenal weight (which was due largely to a reduction in the tion of ACTH (as demonstrated by a fall in adrenal weight, mass of the adrenal cortex). When hypophysectomized rats Table 2), but provided the tissues with more than adequate were treated with ACTH, PNMT activity was largely restored amounts of glucocorticoid. PNMT activity did not fall. Next, with 6 days (Table 1). No other pituitary hormone restored other rats were treated in a like manner with small doses of adrenal PNMT levels after hypophysectomy. metyrapone, a compound that interferes with the synthesis of Most of the fall in adrenal PNMT activity occurred within I adrenal glucocorticoids by inhibiting 11-f3oxidation. Animals week of hypophysectomy; by this time, epinephrine-forming so treated developed adrenal hypertrophy (Table 2), indicating ability was reduced in the majority of rats to 15% to 20% of that ACTH secretion from the pituitary had indeed been ele- normal. Subsequently, PNMT activity decreased much more vated. However, epinephrine-forming ability did not show a slowly, reaching 10% of control levels 10 weeks after hypoph- parallel rise. ysectomy. The content of epinephrine in the adrenals declined On the basis of the above studies, it could be anticipated that significantly within 1 week of hypophysectomy; it continued to glucocorticoids would share with ACTH the ability to elevate fall for as long as animals could be maintained, reaching levels PNMT activity in the hypophysectomized animal, if provided of about 25% to 35% of normal after 10 weeks."This slow fall in adrenal epinephrine levels was thought to be consistent with the epinephrine's known low rate of turnover in adrenals (ie, 7 Table 2. Effects of Dexamethasone and Metyrapone on PNMT to 14"days). Adrenal norepinephrine rose somewhat following Activity in Intact Rats hypophysectomy, as might be anticipated. The adrenals of Adrenal Weight chronically hypophysectomizeddogs were subsequently shown Treatment. (mg/pair) PNMT (,.mollpair) not only to contain subnormal epinephrine levels and PNMT Control 54:!: 1.2 5.12 :!: 0.47 activities, but also to secrete subnormal quantities of epineph- Dexamethasone 35:!: 1.3t 6.02 :!: 0.44 rine basally or when made hypoglycemic by injecting insulin.9 Metyrapone 60 :!: 2.2t 5.12:!: 0.47 Normal epinephrine secretion was restored in this species by *Rats were given 1 mg of the drug intraperitoneally for 6 days and physiologic doses of ACTH. but not by physiologic doses of killed on the 7th day. glucocorticoids (which, as discussed below, perfuse the me- t P < .001 v control. STRESS AND EPINEPHRINE SYNTHESIS 13 in adequate doses. And when hypophysectomized rats were in the general circulation) and on adrenal PNMT activity given the highly potent drug dexamethasone, it, like ACTH, (which, we hypothesized, depends on intra-adrenal blood lev- fully restored epinephrine-forming activity (Table 1). els). It was observed that a dose of ACTH that was equipotent These data were thought to indicate that: (1) adrenal PNMT with a certain amount of hydrocortisone in depressing splenic activity is largely confined to the medulla; and (2) epinephrine weight was fully 100 times more potent in inducing PNMT. synthesis is controlled by the availability of ACTH, but indi- Thus, maintaining the "normal" rate of epinephrine synthesis rectly, via a route that involves the secretion of glucocorticoids requires that the adrenal medulla be surrounded by, and receive from the adrenal cortex. Natural and synthetic steroids that portal venous blood from, a normally functioning adrenal cor- lacked potency as glucocorticoids were found to have no effect tex. Hence, the location of the mammalian medulla within the on PNMT.IO cortex is an important factor in controlling epinephrine syn- Glucocorticoids added to adrenal homogenates from hy- thesis. pophysectomized rats did not stimulate PNMT activity, and large doses actually inhibited this activity. This suggested that CONTROL OF EPINEPHRINE SYNTHESIS WITHIN these hormones were affecting PNMT activity by controlling ADRENAL AND BRAIN AMONG STRESSED ANIMALS the rate of formation (or destruction) of the enzyme protein. To test this hypothesis, the ability of dexamethasone to restore To the best of the author's knowledge, only fragmentary data PNMT activity in hypophysectomized rats was measured in exist on the effects of chronic stress on epinephrine synthesis animals pretreated with actinomycin D or puromycin. Neither and secretion in normal human subjects (although it is well agent had a direct effect on PNMT activity, but both blocked known, of course, that acute stress causes the liberation of large the anticipated rise in epinephrine-forming ability induced by amounts of epinephrine). This is surprising, and should be dexamethasone. This suggested that glucocorticoids secreted rectified, perhaps initially by comparing adrenomedullary from the adrenal cortex act physiologically to induce the syn- PNMT and epinephrine levels of such stressed people with thesis of new PNMT protein within the adrenal medulla, a those in adrenals from people who died without having suffered hypothesis subsequently confirmed by the finding that the hor- a chronic stress. monesalso increasePNMTmRNAlevels.I 1.12 Immobilization stress for 2.S-hour periods has been shown to increase adrenal PNMT activity in rats.12 This effect is best observed after 7 days of treatment, and does not occur in EFFECT OF "REPLACEMENT" DOSES OF hypophysectomized animals. Various regions of rat brain con- GLUCOCORTICOIDS ON EPINEPHRINE-FORMING ABILITY tain low levels of PNMT activity, PNMT mRNA, and epineph- Although PNMT activity was fully restored by low doses of rine,13and the release of intracerebral epinephrine is enhanced ACTH (ie, the doses needed to maintain adrenal weight), the by experimental stress.14The neonatal administration of dexa- amounts of glucocorticoid needed to maintainepinephrine syn- methasone apparently increases brain stem PNMT activity per- thesis seemed unusually high. The usual "replacement" dose of manently, and amplifies the increase in PNMT caused by glucocorticoid in the rat is thought to be about 0.1 to 0.3 mg (of stress.' 5 corticosterone) per day. Doses higher by tenfold had essentially If chronic stress increases the synthesis of epinephrine in, no effect on PNMT. It was necessary to treat rats with 100 to and its secretion from, the human's adrenal medulla, and, as a 300 times the replacement dose (or its equivalent in dexameth- consequence, if it also reduces norepinephrine levels in, and asone) 10 produce a significant restoration of PNMT.'O secretion from, the medulla, what might be the functional Since the effect of ACTH on PNMT had been shown to be consequences of these changes? Epinephrine can markedly indirect, this great difference between ACTH and glucocorti- increase blood glucose levels (by stimulating hepatic glycogen- coid doses suggested that PNMT activity depends not on the olysis and decreasing peripheral glucose uptake). This \vould levels of glucocorticoid in the general circulation, but on levels increase the availability of glucose to the brain, but might also within the adrenal gland itself. Ordinarily, adrenal venous exacerbate diabetes. Epinephrine is also more potent than nor- blood (which bathes the adrenal medulla after it has perfused epinephrine in raising body temperature and in increasing heart the cortex)6contains about 100 times as much glucocorticoid as rate and cardiac output. The expected decrease in adrenomed- peripheral venous blood. When rats are hypophysectomized, ullary norepinephrine secretion would probably be compen- both intra-adrenal and peripheral glucocorticoid levels decline. sated by the stress-induced increase in the release of norepi- Treatment of hypophysectomized rats with "replacement" nephrine from sympathetic nerve terminals. Plasma levels doses of glucocorticoids restores normal hormone levels in the probably do not fall but may actually rise, increasing peripheral general circulation but does not restore intra-adrenal levels. resistance and raising blood pressure. If the pituitary was un- However, when such animals are given ACTH, intra-adrenal able to secrete additional ACTH in response to the stress, this corticoid levels do return to normal. Thus it was concluded that might, by impairing the rise in epinephrine synthesis and se- the restoration of PNMT by customary doses of ACTH results cretion, compromise the ability of the brain and heart, and from the selective increases they produce in intra-adrenal glu- perhaps other organs, to respond appropriately. cocorticoid levels. In order to test this hypothesis, hypophysec- A stress-induced increase in brain epinephrine synthesis tomized rats were treated with several different doses of ACTH might be expected to affect the central regulation of cardiovas- or of hydrocortisone, and the effects of these hormones were cular function. The exact consequences of this putative increase measured on splenic weight (which depends on corticoid levels await characterization. 14 RICHARD J. WURTMAN
REFERENCES
1. Wurtman Rl, Axelrod 1: Adrenaline synthesis: Control by the of epinephrine in response to insulin among hypophysectomized dogs. pituitary gland and adrenal glucocorticoids. Science ]50: 1464-1465, Proc Natl Acad Sci USA 61:522-528, 1968 1965 10. Wurtman RJ: Control of epinephrine synthesis in the adrenal 2. Pohorecky LA, Wurtman RJ: Adrenocortica] control of epineph- medulla by the adrenal corte)c Hormonal specificity and dose-response rine synthesis. Pharmaco] Rev 23: 1-35, ] 971 characteristics. Endocrinology 79:608-614, 1966 3. Green WL, Ingbar SH: Decreased corticotropin reserve as iso- II. Stachowiak MK, Rigual RJ, Lee PHK, et al: Regulation of lated pituitary defect. Arch Intern Med 108:945-952, ]961 tyrosine hydroxylase and phenylethanolamine N-methyltransferase 4. Luft R, von Euler US: Effect of insulin hypoglycemia on urinary mRNA levels in the sympathoadrenal system by the pituitary-adreno- cortical axis. Brain Res 427:275-286, 1988 excretion of adrenaline and noradrenaline in man after hypophysec- 12. Cahill AL, Eertmoed AL, Mangoura D, et al: Differentia] reg- tomy.l Clin Endocrinol Metab ]6:1017-1025, 1956 u]ation of pheny]ethanolamine N-methy]transferase expression in two 5. Hung W, Migeon Cl: Hypoglycemia in a two-year old by with distinct subpopulations in bovine chromaffin cells. 1 Neurochem 67: adrenocorticotropic hormone (ACfH) deficiency (probably isolated) 1217-1224, 1996 and adrenal medullary unresponsiveness to insulin-induced hypogly- 13. Pohorecky LA, Zigmond M, Karten H, et a]: Enzymatic con- cemia. 1 Clin Endocrinol Metab 28:146-152, 1968 version of norepinephrine to epinephrine by the brain. 1 Pharmacol Exp 6. Coupland RE: On the morphology and adrenaline noradrenaline Ther 165:190-195, 1969 content of chromaffin tissue. 1 Endocrinol 9: 194-203, 1953 14. Sauter AM, Baba Y, Stone EA, et al: Effect of stress and of 7. Kirshner N: The formation of adrenaline from noradrenaline. phenylethanolamine-N-methyltransferase inhibition on central norepi- Biochem Biophys Acta 24:658-659, 1959 nephrine and epinephrine levels. Brain Res 144:415-419, 1978 8. Axelrod 1: Purification and properties of phenylethanolamine-N- 15. Turner BB, Katz RJ, Carroll Bl: Neonatal corticosteroid perma- methyl transferase. 1 BioI Chern 237: 1657-1660, 1962 nently alters brain activity of epinephrine-synthesizing enzyme in 9. Wurtman RJ, Casper A, Pohorecky LA, et al: Impaired secretion stressed rats. Brain Res 166:426-430, 1979